FULL-LENGTH ORIGINAL RESEARCH
Frontal and thalamic changes of GABA concentration indicate dysfunction of thalamofrontal networks in juvenile myoclonic epilepsy *Elke Hattingen, *Christian L€ uckerath, *Stefanie Pellikan, *Dmitri Vronski, †Christine Roth, †Susanne Knake, ‡Matthias Kieslich, and *Ulrich Pilatus Epilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
SUMMARY
Elke Hattingen is a German neuroradiologist and expert on metabolic MR imaging Goethe University Frankfurt.
Objective: Juvenile myoclonic epilepsy (JME) has been considered to be a frontal variant of thalamocortical network dysfunction in epilepsy. Changes of c-aminobutyric acid (GABA)ergic neurotransmission may play a key role in this dysfunction. Magnetic resonance spectroscopy (MRS) is the only noninvasive method to measure GABA concentrations in different brain regions. We measured GABA and other metabolite concentrations in the thalamus and frontal lobe of patients with JME. Methods: A specific protocol was used for determining GABA concentrations in the thalamus, frontal lobe, and motor cortex contralateral to the handedness in 15 patients with JME and 15 age-matched controls. In addition, we measured concentrations of glutamate and glutamine, N-acetyl-aspartate (NAA), myoinositol, creatine, and choline using MRS with short echo time. JME-related concentration changes were analyzed comparing patients to controls, also considering potential effects of antiepileptic drugs. Results: In patients with JME, GABA and NAA were reduced in the thalamus (p = 0.03 and p = 0.02), whereas frontal GABA and glutamine were elevated (p = 0.046 and p = 0.03). MRS revealed reduced NAA in the thalamic gray matter contralateral to the handedness (p = 0.04 each). These changes were found consistently in patients treated with new antiepileptic drugs and with valproate, although the extent of metabolic changes differed between these treatments. Significance: Decreased thalamic and increased frontal GABA suggest a dysfunction of GABAergic neurotransmission in these brain regions of patients with JME. The NAA decrease in the gray matter of the thalamus may hint to a damage of GABAergic neurons, whereas frontal increase of GABA and its precursor glutamine may reflect increased density in GABAergic neurons due to subtle cortical disorganization in the thalamofrontal network. KEY WORDS: Juvenile myoclonic epilepsy, GABA, MR spectroscopy, Thalamofrontal.
Accepted April 12, 2014; Early View publication June 5, 2014. *Department of Neuroradiology, Goethe University Frankfurt, Frankfurt, Germany; †Centrum of Epilepsy Hessen, Philipps-Universit€at Marburg, Marburg, Germany; and ‡Department of Paediatric Neurology, Goethe University Frankfurt, Frankfurt, Germany Address correspondence to Elke Hattingen, Department of Neuroradiology, Goethe University Frankfurt, Schleusenweg 2-16, Frankfurt 60528, Germany. E-mail: [email protected] Wiley Periodicals, Inc. © 2014 International League Against Epilepsy
Juvenile myoclonic epilepsy (JME) is characterized by awakening myoclonus, often associated with tonic–clonic seizures, absences, and neuropsychological frontal lobe pathology.1,2 Abnormal oscillations of the thalamocortical networks are considered to trigger myoclonic seizures in JME.3 The thalamus amplifies and synchronizes thalamocortical rhythms, which involve pyramidal cells of the motor cortex.4 This thalamocortical network is reciprocally connected by excitatory glutamatergic and inhibitory c-aminobutyric acid (GABA)ergic projections.5 Congenital
1030
1031 Frontal and Thalamic GABA Concentration in JME alterations of GABAA receptors play a key role in JME, and changes in GABAA receptor were found in positron emission tomography (PET) imaging.6,7 Apart from the receptor level, there is increasing evidence that the network between the frontal lobe and the thalamus is structurally disturbed during early brain development in patients with JME.8 GABA is not only functionally but also metabolically linked with its excitatory counterpart glutamate,9 as glutamine is the precursor of both. GABA and glutamate, once released from neurons, are taken up into astrocytes, which convert glutamate into glutamine. Impairments of these glial-neuronal interactions result in seizures.10 High-field proton magnetic resonance spectroscopy (1H MRS) allows noninvasive quantification of these neurotransmitters. However, glutamine and glutamate are not easily quantifiable at field strengths below 4 Tesla when their overlapping peaks start to separate and GABA measurement requires additional editing. In contrast, the neuronal marker N-acetyl aspartate-glutamate (NAA) is quantifiable even at lower field strengths, and reduced NAA was observed in brain structures that are involved in epilepsy.11–15 Only few studies measured in vivo concentrations of GABA or glutamate and glutamine separately (Glx as sum of both, Glx) in patients with JME.16–18 Petroff et al.16,17 found decreased GABA concentrations in the occipital brain of JME patients, whereas Simister et al.14,18 reported increased GABA levels in the occipital lobe of 15 patients with idiopathic generalized epilepsy (IGE), including two patients with JME. Levels of Glx were increased in the frontal lobe, the insula, and striatum of patients with JME, raising the question whether this increase refers to glutamine or glutamate or both. Herein we present data on metabolite concentrations in brain structures that are most relevant for JME (thalamus, frontal lobe, and motor cortex) focusing on changes of GABA, glutamate, glutamine, and NAA. In addition to 1H MRS with short echo time (TE) to measure glutamate and glutamate separately, a specific MR spectroscopic GABA A
C
editing sequence was employed. Because metabolite concentrations can also be changed by the medication used for treatment,19–24 the study was also designed to obtain information of those effects, as well hemispheric and tissue differences.
Materials and Methods Study population This prospective noninterventional study was approved by the institutional review board (Ethics committee at the University Hospital Marburg, 150/11), and written informed consent was obtained from each participant prior to inclusion. Patients were enrolled by an epilepsy center, and diagnosis of JME was confirmed by electroencephalography (EEG) and semiology. Diagnosis of JME was made according to the criteria of the 2012 International Consensus Criteria by an epilepsy expert,25 defined by myoclonic jerks without loss of consciousness on awakening, normal interictal EEG with ictal high-amplitude generalized epileptiform discharges (class I), and otherwise normal intelligence and normal neurologic examination. Patients and healthy age-matched controls had a normal morphologic cranial magnetic resonance imaging (MRI) study. Sixteen JME patients and 18 healthy subjects were enrolled between 2011 and 2012. None of the patients had a seizure 24 h before the MR examination. MR spectroscopy All spectroscopic examinations were performed at 3 Tesla. The entire protocol, which was initially designed to obtain 1H- and 31P-MRS data from brain areas with potential pathologic relevance for JME, was divided into two sessions (MR parameter details are indicated in the supporting information): 1 Single-voxel spectroscopy (SVS) to measure 1H MRS detectable metabolite concentrations in the thalamus, the frontal lobe, and the motor cortex of the hemisphere D
B
Figure 1. Single voxel position in the frontal lobe (A) and thalamus (B), and representative spectra for analysis of short echo time (TE) (C) and GABA-edited (D) signals. The dotted red line in C shows the result from the LCModel fit. Intensity adjusted basis spectra for MI (blue) and Gln (orange), as obtained by the fit, are also included. GABA-editing by MEGA-PRESS (TE = 70 msec) is shown in (D). The black line represents the difference between the blue (editing pulse at 4.1 ppm) and green signal (editing pulse at 1.9 ppm). The fitting procedure quantifies the GABA signal at 3 ppm (dotted red line). Epilepsia ILAE Epilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
1032 E. Hattingen et al. contralateral to the handedness (Fig. 1). The SVS data included a short TE (30 msec) point resolved selective spectroscopy (PRESS) sequence to measure the main 1H MRS detectable metabolites (see below) and a MescherGarwood (MEGA)-PRESS sequence at a TE of 70 msec, dedicated to the detection of GABA by using a a pair of selective excitation pulses at 1.9 ppm.27 2 In a separate study short TE (30 msec) 1H MR spectroscopic imaging (MRSI) was performed. A weighted circular phase-encoding scheme was used with a 24 9 24 matrix and 240 9 240 mm2 field of view. The volume of interest was selected by a combination of pointresolved selective spectroscopy and outer volume suppression. Before Fourier transformation, the matrix was extrapolated to 48 9 48, resulting in a 5 9 5 mm2 inplane grid size. To obtain metabolite concentrations in gray matter (GM) or white matter (WM), co-registration with segmented anatomic images for GM and WM was performed to allow manual selection of the small MRSI voxels with only GM and WM (see Figure S1 in Supporting Information). The slices were coronal oriented to measure brain metabolites in GM and WM of both hemispheres including the thalamus and the motor cortex. Data processing MRS data Short TE SVS and MRSI spectra were analyzed quantifying the main metabolites NAA (N-acetyl-aspartate/N-acetylaspartate-glutamate), Cr (creatine/phosphocreatine), Cho (choline-containing compounds), and MI (myoinositol). The larger SVS voxel > 8 ml together with high spectral resolution at 3 T allowed quantification of glutamine and glutamate separately, whereas Glx as sum of glutamine and glutamate was quantified for the much smaller MRSI voxels (0.30 ml). No GABA concentrations were available from MRSI spectra, since detection and quantification of GABA requires special sequences (e.g. MEGA-PRESS).16–18,26 The GABA-edited data (Fig. 1D) from MEGA-PRESS spectra were analyzed with the tool jMRUI while the shortTE 1H MRS spectra were analyzed with the software tool LCModel (http://s-provencher.com). Partial volume for GM, WM, and cerebrospinal fluid (CSF) was obtained from the segmented anatomic images. Further details of data acquisition and processing are provided in the Supporting Information. Statistics Statistical analysis was performed with STATISTICA (version 7.1; StatSoft, Tulsa, OK, U.S.A.). For all tests, p < 0.05 was considered to be statistically significant. Comparing JME patients to controls Concentration differences for GABA and NAA between patients and control subjects were tested with analysis of Epilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
variance (ANOVA) using the coil loading as covariate. Different target regions were considered as repetitive measurements. Contrast analysis in ANOVA was performed for each metabolite and target region. Effects of antiepileptic drugs Previous data revealed that antiepileptic drugs (AED) have an effect on brain metabolites in 1H MRS. It has been shown that valproate induces changes in cerebral myoinositol and glutamine concentrations, whereas new AEDs may increase GABA levels in patients with epilepsy.21–25,29 Therefore, we divided the patients into two groups: one group included patients with valproate treatment, whereas the other group covered all patients who received any medication but no valproate. Differences in myoinositol and glutamine between these two groups were tested using ANOVA.
Results Study population For three controls and one class I JME patient treated with lamotrigine, no MRSI was performed due to subject’s cancellation. These subjects were not included in the analysis. The remaining 30 subjects had a median age for patients and controls of 25 years. Age of the patients was 16–36 years, except of one 54-year-old patient; age range of controls was 16–38 years. The control group consisted of seven females and eight males, whereas the JME group included 10 females and five males. Clinical data of the JME patients are given in Table 1. Typical semiology with myoclonic jerks occurring on or after awakening without loss of consciousness was reported by all patients. Thirteen of the 15 patients had typical myoclonic seizures as first symptom of JME, which manifested at the age of 10–19 years, whereas in one patient, myoclonic jerks occurred together with absences and follow-up EEG was typical for juvenile absence epilepsy (Table 1). In another patient the myoclonus was associated with awakening tonic–clonic seizure early after onset, fulfilling the class 2 criteria according the International Consensus.25 The remaining 13 patients with JME semiology and typical EEG findings fulfilled the class 1 criteria according the International Consensus.25 Seven patients had generalized tonic– clonic seizures and absence seizures over the years of illness. Two patients had generalized tonic–clonic and two patients had absence seizures. All but one patient had antiepileptic therapy, which consisted of valproate, levetiracetam, lamotrigine, topiramate, or the combination of those. To ensure that the uncertain case of JME or gender differences did not influence the major effects of neurotransmitter changes, we performed a second statistical analysis on all SVS data, only, matching the groups for age and gender (10 female and five male, mean age 26) and replacing the data of the uncertain case by those from the other class I JME patient.
1033 Frontal and Thalamic GABA Concentration in JME Table 1. Clinical characteristics of the patients and antiepileptic drugs Patient
Age (y), gender
Age at onset (y)
Disease duration (y)
JME class according26
Seizure characteristics
AEDs
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
35, m 26, f 16, f 36, m 31, f 25, m 25, m 16, m 32, f 20, f 16, f 54, f 17, f 21, f 20, f
17 18 13 18 11 11 19 14 13 13 10 10 11 16 16
18 8 3 18 20 14 6 3 19 7 6 44 6 5 4
1 1 1 1 1 1 1 1 1 1 1 1
M, GTC, A M M M, GTC, A M, GTC, A M, GTC M M, GTC, A M, GTC, A M, A M, GTC, A M M, A M, GTC M, GTC, A
VPA, LTG LTG VPA LEV LEV VPA VPA VPA VPA LEV LEV, LTG None LEV VPA, LTG VPA, TPM
a
1 2
Y, years; m, male, f, female; AED, antiepileptic drugs; M, myoclonia; GTC, generalized tonic–clonic seizure; A, absence; VPT, valproate; LEV, levetiracetam; LTG, lamotrigine; TPM, topiramate. a Included as JME. EEG showed generalized 3/s spike-slow wave discharges intensified by photo stimulation characterizing juvenile absence epilepsy.
Metabolite concentrations Concentrations of the metabolites GABA, NAA, and glutamine as obtained from SVS magnetic resonance spectroscopy (MRS) of the hemisphere contralateral to the handedness differed for the selected brain areas between healthy control subjects and patients (Fig. 2) showing: 1 Increase in GABA concentrations in the frontal region (p = 0.046) and decreased GABA concentrations in the thalamus (p = 0.03). 2 Decrease in NAA concentrations in the thalamus (p = 0.02). 3 Increase of glutamine concentration in the frontal area (p = 0.03). These results were confirmed by the additional analysis of the second cohort with exactly matched gender and age. Table 2 also shows metabolites of the patient group dichotomized into a group treated with valproate and a group without valproate treatment. These data should exclude that above-mentioned significant metabolite changes reflect an effect of AED instead of being disease related. Although the statistical power of a comparison between each patient group and controls is rather low, the above-mentioned significant changes in JME patients were found equally in each subgroup. The previously reported effects of valproate on metabolites were also found in terms of a consistent increase in glutamine and a decrease in myoinositol when comparing the two patient groups, and ANOVA revealed significance (p = 0.045); contrast analysis in ANOVA confirmed increased glutamine in the thalamus. Although the SVS data provided only metabolite concentration changes in the target region of the hemisphere contralateral to the handedness, the MRSI data also included
information for the target region ipsilateral to the handedness and allowed for discrimination between GM and WM (Supporting Information). However, concentrations of GABA and glutamine were not obtained with these data, and the frontal region was not covered by the coronal slice. An intra-individual comparison for control subjects showed significantly higher NAA concentration in the GM of the thalamus contralateral to the handedness compared to ipsilateral (p = 0.006). Comparing patients to controls revealed significantly decreased concentrations of NAA (p = 0.04) and myoinositol (p = 0.002) for this tissue, confirming the SVS data (Appendix S1). Both changes were also visible in the hemisphere ipsilateral to the handedness, but they did not reach significance. Furthermore, choline was significantly decreased in the thalamic GM of the hemisphere contralateral to the handedness (p = 0.04). Dichotomizing into the two groups with different medication shows the most prominent decrease of myoinositol and NAA in the thalamic GM of patients treated with valproate.
Discussion This study focuses on potentially affected brain areas when measuring GABA, glutamine, and glutamate concentration in JME patients. The primary findings were reduced concentrations of neuronal metabolites GABA and NAA in the thalamus, whereas GABA and glutamine were increased in the frontal lobe. Although these results are based on SVS data targeting the contralateral to handedness side, the MRSI data confirmed reduced NAA concentrations in the thalamus contralateral to the handedness in JME patients, also pointing to more prominent changes in the gray matter. Petroff et al.16 found a decrease of GABA concentration in Epilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
1034 E. Hattingen et al.
Figure 2. Metabolites from the glutamate/GABA-glutamine cycle and the neuronal marker NAA as obtained from SVS of contralateral to the handedness hemisphere for healthy subjects (Prob) and patients (Pat). The box plots represent mean values with standard errors; standard deviation is marked by whiskers. Epilepsia ILAE Epilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
the occipital brain region of JME patients, but they did not make an attempt to localize GABA changes in different brain areas. Nevertheless, these occipital measurements may have included thalamic regions, since SVS was performed with an 8-cm surface coil and a large voxel size of 14 ml. In contrast, Simister et al.17,18 used localized PRESS with double quantum filter to measure GABA in the frontal and occipital lobe of patients with IGE, detecting GABA increase only in the occipital area. Only part of these IGE patients had the diagnosis of JME (2 of 15 patients for occipital measurements and 7 of 21 patients for frontal measurements). In contrast, our study included JME patients meeting the internationally proposed class I (n = 14) or class II (n = 1) JME criteria.25 Apart from the methodologic differences, these results suggest that regional differences of altered GABA concentration are responsible for different phenotypes of IGE seizures. In a PET study investigating a selective antagonist of GABAA receptors, [11C]-flumazenil (FMZ), Koepp et al.7 showed an increased binding in the cerebral cortex and thalamus of 10 IGE patients; in 5 JME patients there was an accentuation of the dorsolateral prefrontal areas when compared to healthy controls. Because a mutation in the a1 subunit of the GABAA receptor was found in a family with autosomal JME,6 one key in the pathogenesis of JME was attributed to the receptor level. In vitro studies revealed that the majority of the mutations result in a reduction of GABA-activated chloride currents in recombinant receptors.27 GABA is the main inhibitory transmitter in the adult brain; therefore, it was suggested that the lack of GABA-activated chlorine currents caused abnormal excitability of cortical neurons, thereby resulting in myoclonic or general seizures. Apart from loss of function, there is increasing evidence that epilepsy-causing mutations in GABAA receptors might also be associated with abnormal development of neuronal networks, which is anticipated as one of the critical mechanisms leading to the disease. During brain development, GABA seems to act as a excitatory neurotransmitter, playing a key role in various aspects of the maturation process, such as neuronal migration.28 Furthermore, it has been found recently that other mutations associated with JME are deeply involved in cerebral corticogenesis: The myoclonin1/EFHC1 microtubule-associated protein is involved in cell division and radial migration during cortical development, and mutations in the EFHC1 gene caused a marked disturbance of cortical development in vivo.8 BRD2, the locus for another JME-inducing gene (EJM1), is a putative transcriptional regulator from a gene family that is expressed during cerebral development.5 Abnormal brain development in JME is supported by structural MRI studies. Voxel-based volumetry and diffusion tensor imaging reveal changes predominantly in the prefrontal cortex involving the gray and underlying white matter.29–31 Furthermore, the connectivity of different brain regions was found to be
1035 Frontal and Thalamic GABA Concentration in JME Table 2. Values with standard deviation (italic) of all metabolites measured with single voxel spectroscopy (SVS). Significant differences between patients and control subjects are marked in bold. Values represent concentrations in mM Patients Region Motor area
Frontal
Thalamus
Control
Metabolite Cr MI Cho NAA Glu Gln GABA Cr MI Cho NAA Glu Gln GABA Cr MI Cho NAA Glu Gln GABA
7.09 4.14 1.30 11.10 7.06 4.21 1.97 5.56 3.51 1.36 8.79 5.25 2.20 3.09 6.44 3.50 1.53 9.56 5.66 2.67 4.29
Total 0.86 0.79 0.17 1.57 1.43 2.05 0.46 1.27 0.43 0.31 1.39 1.33 1.29 0.63 0.76 0.54 0.16 1.02 0.76 1.17 0.50
7.07 3.83 1.39 10.64 6.74 4.17 2.04 5.60 3.31 1.42 8.71 5.24 3.16 3.56 6.39 3.52 1.55 8.95 5.39 3.17 3.85
affected, either by means of a decrease to frontal areas or by an increase to the motor pathways.32 Although all studies reported structural changes in patients with JME, the areas and the direction of changes still differed between these studies.29,30,33 Increased frontal concentration levels of GABA and of the neurotransmitter precursor glutamine found in this study corroborate the findings of an involvement of the frontal cortex. MRS measures the total amount of GABA that is available. Thus, the GABA concentration mainly reflects the fractional volume of GABAergic neurons adjusted by the vesicular and synaptic GABA content of those neurons. Furthermore, metabolic concentrations measured by MRS do not simply represent the neurotransmitter pool, but also reflect the metabolism. Increased GABAA receptor binding as observed for the JME patients in the PET study described previously,7 would be in line with an increased density in GABAergic neurons in the frontal lobe. An increased density in GABAergic neurons also supports the hypothesis that developmental disorganization of the cortex in form of cortical and subcortical dystopic neurons (“cortical microdysgenesis”) may play a key role in JME.7,34 This would also explain why higher frontal gray matter concentration of JME patients were found compared to controls.29,35 The relevance of glutamine (maybe plus glutamate) for JME is supported by previous findings. Simister et al.17,18 found increased levels of Glx (glutamate + glutamine) in the frontal and occipital lobe of patients with IGE. In accordance with our study results, frontal Glx increase was also found
Valproate 1.03 0.76 0.28 1.43 1.14 1.57 0.68 1.21 0.68 0.28 1.34 0.80 1.11 0.64 0.94 0.76 0.23 1.26 1.01 0.80 0.67
6.83 3.54 1.27 10.00 6.35 4.91 1.89 5.17 3.02 1.34 8.01 4.90 3.28 3.42 6.09 3.13 1.47 8.37 5.10 3.54 3.63
Nonvalproate 1.02 0.59 0.34 1.43 0.92 1.78 0.70 1.30 0.48 0.29 1.10 0.66 1.33 0.56 1.06 0.78 0.27 1.35 1.13 0.90 0.71
7.36 4.17 1.51 11.35 7.18 3.31 2.20 6.10 3.65 1.51 9.50 5.64 3.04 3.73 6.72 3.96 1.65 9.61 5.74 2.77 4.10
1.06 0.83 0.12 1.13 1.26 0.68 0.67 0.94 0.78 0.23 1.21 0.79 0.88 0.55 0.71 0.47 0.11 0.78 0.79 0.44 0.58
in the subgroup of seven JME patients.18 Another study found a frontal Glx/Cr increase together with NAA/Cr decrease in JME patients with personality disorders, supporting the hypothesis that typical personality changes of JME patients are related to the frontal lobe dysfunction.15 The decrease of GABA and NAA in the thalamus may be attributed to a reduction of the GABAergic neurons. NAA is formed in the neuronal mitochondria, and hereby the NAA decrease marks neuronal dysfunction or neuronal loss. NAA decrease in the thalamus of JME patients has been reported previously.12–14 Inconsistent results were found in voxel-based morphometry studies concerning neuronal damage in the thalamus of JME patients.30,35,36 However, some studies hint to a regional abnormality in the anterior and medial parts of the thalamus, which are connected to the frontal lobe, supporting the role of thalamofrontal network dysfunction in the epileptogenesis of JME. In our study, concentration decreases of NAA, but also of Cho and MI were most prominent in the GM of the thalamus contralateral to the handiness, reflecting a higher vulnerability of this tissue in JME. Language specialization of the thalamus contralateral to the handedness and functional demand according to the handedness may enhance metabolic changes in the thalamus.37 The hemispheric difference can also explain why Savic et al.38 did not find a significant NAA decrease in the right thalamus, which is ipsilateral to the handedness in the majority of humans. However, it remains unclear whether these changes are of secondEpilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
1036 E. Hattingen et al. ary degenerative nature or if they are part of a cerebral development failure. An experimental study in heterozygous BRD2 mice hints to a development failure.5 It is worth mentioning that NAA concentration in the thalamus contralateral to the handedness is higher in healthy subjects (see Appendix S1).39 All JME patients except one were receiving antiepileptic treatment. Because interruption of this treatment is contraindicated, potential drug effects on brain metabolite concentrations have to be considered. In particular, it should be excluded that the above-mentioned changes in GABA, glutamine, and NAA are drug-induced. According to previously published data, valproate should not change the GABA and NAA concentrations, whereas other medications might increase GABA.19,21–24 On the other hand, valproate increases cerebral concentrations of glutamine and decreases the myoinositol concentration.22,23 However, dichotomized patient data reveal that both, thalamic GABA decrease as well as frontal increase of GABA and glutamine, should be related to the disease, given that these changes were found equally in both groups. Of note, the increased concentration of neuronal metabolites NAA, GABA, and glutamate in the frontal area and the motor cortex were more prominent under therapy with newer AED compared with the valproate treatment, which might reflect a direct effect on cerebral neuronal metabolism for these drugs.21,24 Glutamate and GABA are interconnected via the glutamate/ GABA-glutamine cycle and NAA is discussed as a reservoir for glutamate.9,40 These findings also support the hypothesis that some drugs interacting with neurotransmitter systems may have a neurotropic effect.20 In conclusion, changes of GABA concentration in the thalamus and frontal lobe as key regions for seizure triggering may be essential in JME. The decrease of the neuronal marker NAA in the gray matter of the thalamus implicates a damage of GABAergic neurons. The increase of GABA and its precursor glutamine hints to inborn errors of the thalamofrontal neuronal network, either on the receptor level or in the form of failures in corticogenesis. Although GABA concentrations are affected by some AEDs, we can exclude a simple treatment effect. Larger MRS trials should approve the generalizability of our study results by analyzing neurotransmitter levels in JME patients before and after initiating AED treatment.
Funding There was no external funding for the study.
Disclosure All authors declare that none of the authors has any conflict of interest to disclose. All authors confirm that they have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines. Epilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
References 1. Canevini MP, Mai R, Di Marco C, et al. Juvenile myoclonic epilepsy of Janz: clinical observations in 60 patients. Seizure 1991;1:291–298. 2. Devinsky O, Gershengorn J, Brown E, et al. Frontal functions in juvenile myoclonic epilepsy. Neuropsychiatry Neuropsychol Behav Neurol 1997;10:243–246. 3. O’Muircheartaigh J, Vollmar C, Barker GJ, et al. Abnormal thalamocortical structural and functional connectivity in juvenile myoclonic epilepsy. Brain 2012;135:3635–3644. 4. Gotman J, Grova C, Bagshaw A, et al. Generalized epileptic discharges show thalamocortical activation and suspension of the default state of the brain. Proc Natl Acad Sci USA 2005;102:15236– 15240. 5. Velısek L, Shang E, Velıskova J, et al. GABAergic neuron deficit as an idiopathic generalized epilepsy mechanism: the role of BRD2 haploinsufficiency in juvenile myoclonic epilepsy. PLoS ONE 2011;6: e23656. 6. Cossette P, Liu L, Brisebois K, et al. Mutation of GABRA1 in an autosomal dominant form of juvenile myoclonic epilepsy. Nat Genet 2002;31:184–189. 7. Koepp MJ. Juvenile myoclonic epilepsy– a generalized epilepsy syndrome? Acta Neurol Scand Suppl 2005;181:57–62. 8. de Nijs L, Leon C, Nguyen L, et al. EFHC1 interacts with microtubules to regulate cell division and cortical development. Nat Neurosci 2009;12:1266–1274. 9. Bak LK, Schousboe A, Waagepetersen HS. The glutamate/GABAglutamine cycle: aspects of transport, neurotransmitter homeostasis and ammonia transfer. J Neurochem 2006;98:641–653. 10. Eid T, Ghosh A, Wang Y, et al. Recurrent seizures and brain pathology after inhibition of glutamine synthetase in the hippocampus in rats. Brain 2008;131:2061–2070. 11. Savic I, Lekvall A, Greitz D, et al. MR spectroscopy shows reduced frontal lobe concentrations of N-acetyl aspartate in patients with juvenile myoclonic epilepsy. Epilepsia 2000;41:290–296. 12. Mory S, Li L, Guerreiro C, et al. Thalamic dysfunction in juvenile myoclonic epilepsy: a proton MRS study. Epilepsia 2003;44:1402– 1405. 13. Haki C, G€um€ustas OG, Bora I, et al. Proton magnetic resonance spectroscopy study of bilateral thalamus in juvenile myoclonic epilepsy. Seizure 2007;16:287–295. 14. Lin K, Carrete H Jr, Lin J, et al. Magnetic resonance spectroscopy reveals an epileptic network in juvenile myoclonic epilepsy. Epilepsia 2009;50:1191–1200. 15. de Araujo Filho GM, Lin K, Lin J, et al. Are personality traits of juvenile myoclonic epilepsy related to frontal lobe dysfunctions? A proton MRS study. Epilepsia 2009;50:1201–1209. 16. Petroff OA, Hyder F, Rothman DL, et al. Homocarnosine and seizure control in juvenile myoclonic epilepsy and complex partial seizures. Neurology 2001;56:709–715. 17. Simister RJ, McLean MA, Barker GJ, et al. A proton magnetic resonance spectroscopy study of metabolites in the occipital lobes in epilepsy. Epilepsia 2003;44:550–558. 18. Simister RJ, McLean MA, Barker GJ, et al. Proton MRS reveals frontal lobe metabolite abnormalities in idiopathic generalized epilepsy. Neurology 2003;61:897–902. 19. Petroff OA, Rothman DL, Behar KL, et al. Effects of Valproate and other antiepileptic drugs on brain glutamate, glutamine, and GABA in patients with refractory complex partial seizures. Seizure 1999;8:120– 127. 20. Moore GJ, Bebchuk JM, Wilds IB, et al. Lithium-induced increase in human brain grey matter. Lancet. 2000 Oct 7;356(9237):1241–1242. Erratum in: Lancet 2000;356:2104. 21. Kuzniecky R, Ho S, Pan J, et al. Modulation of cerebral GABA by topiramate, lamotrigine, and gabapentin in healthy adults. Neurology 2002;58:368–372. 22. Simister RJ, McLean MA, Barker GJ, et al. The effect of sodium valproate on proton MRS visible neurochemical concentrations. Epilepsy Res 2007;74:215–219. 23. Garcia M, Huppertz HJ, Ziyeh S, et al. Valproate-induced metabolic changes in patients with epilepsy: assessment with H-MRS. Epilepsia 2009;50:486–492.
1037 Frontal and Thalamic GABA Concentration in JME 24. Doelken MT, Hammen T, Bogner W, et al. Alterations of intracerebral c-aminobutyric acid (GABA) levels by titration with levetiracetam in patients with focal epilepsies. Epilepsia 2010;51:1477–1482. 25. Kasteleijn-Nolst Trenite DG, Schmitz B, Janz D, et al. Consensus on diagnosis and management of JME: from founder’s observations to current trends. Epilepsy Behav 2013;28(Suppl. 1):87–90. 26. Mescher M, Merkle H, Kirsch J, et al. Simultaneous in vivo spectral editing and water suppression. NMR Biomed 1998;11:266–272. 27. Macdonald RL, Gallagher MJ, Feng HJ, et al. GABA(A) receptor epilepsy mutations. Biochem Pharmacol 2004;68:1497–1506. Review. 28. Ben-Ari Y, Khalilov I, Kahle KT, et al. The GABA excitatory/ inhibitory shift in brain maturation and neurological disorders. Neuroscientist 2012;18:467–486. 29. Woermann FG, Free SL, Koepp MJ, et al. Abnormal cerebral structure in juvenile myoclonic epilepsy demonstrated with voxel-based analysis of MRI. Brain 1999;122:2101–2108. 30. Kim JH, Lee JK, Koh SB, et al. Regional grey matter abnormalities in juvenile myoclonic epilepsy: a voxel-based morphometry study. Neuroimage 2007;s37:1132–7113. 31. Deppe M, Kellinghaus C, Duning T, et al. Nerve fiber impairment of anterior thalamocortical circuitry in juvenile myoclonic epilepsy. Neurology 2008;71:1981–1985. 32. Vollmar C, O’Muircheartaigh J, Barker GJ, et al. Motor system hyperconnectivity in juvenile myoclonic epilepsy: a cognitive functional magnetic resonance imaging study. Brain 2011;134:1710– 1719. 33. Liu M, Concha L, Beaulieu C, et al. Distinct white matter abnormalities in different idiopathic generalized epilepsy syndromes. Epilepsia 2011;52:2267–2275.
34. Meencke HJ, Janz D. Neuropathological findings in primary generalized epilepsy: a study of eight cases. Epilepsia 1984;25:8–21. 35. Mory SB, Betting LE, Fernandes PT, et al. Structural abnormalities of the thalamus in juvenile myoclonic epilepsy. Epilepsy Behav 2011;21:407–411. 36. Saini J, Sinha S, Bagepally BS, et al. Subcortical structural abnormalities in juvenile myoclonic epilepsy (JME): MR volumetry and vertex based analysis. Seizure 2013;22:230–235. 37. Eidelberg D, Galaburda AM. Symmetry and asymmetry in the human posterior thalamus. I. Cytoarchitectonic analysis in normal persons. Arch Neurol 1982;39:325–332. 38. Savic I, Ostermann Y, Helms G. MRS shows syndrome differentiated metabolite changes in human-generalized epilepsies. Neuroimage 2004;21:163–172. 39. Nagae-Poetscher LM, Bonekamp D, Barker PB, et al. Asymmetry and gender effect in functionally lateralized cortical regions: a proton MRS imaging study. J Magn Reson Imaging 2004;19:27–33. 40. Clark JF, Doepke A, Filosa JA, et al. N-acetylaspartate as a reservoir for glutamate. Med Hypotheses 2006;67:506–512.
Supporting Information Additional Supporting Information may be found in the online version of this article: Appendix S1. 1H MRSI and SVS in JME.
Epilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
Frontal and thalamic changes of GABA concentration indicate dysfunction of thalamofrontal networks in juvenile myoclonic epilepsy *Elke Hattingen, *Christian L€ uckerath, *Stefanie Pellikan, *Dmitri Vronski, †Christine Roth, †Susanne Knake, ‡Matthias Kieslich, and *Ulrich Pilatus Epilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
SUMMARY
Elke Hattingen is a German neuroradiologist and expert on metabolic MR imaging Goethe University Frankfurt.
Objective: Juvenile myoclonic epilepsy (JME) has been considered to be a frontal variant of thalamocortical network dysfunction in epilepsy. Changes of c-aminobutyric acid (GABA)ergic neurotransmission may play a key role in this dysfunction. Magnetic resonance spectroscopy (MRS) is the only noninvasive method to measure GABA concentrations in different brain regions. We measured GABA and other metabolite concentrations in the thalamus and frontal lobe of patients with JME. Methods: A specific protocol was used for determining GABA concentrations in the thalamus, frontal lobe, and motor cortex contralateral to the handedness in 15 patients with JME and 15 age-matched controls. In addition, we measured concentrations of glutamate and glutamine, N-acetyl-aspartate (NAA), myoinositol, creatine, and choline using MRS with short echo time. JME-related concentration changes were analyzed comparing patients to controls, also considering potential effects of antiepileptic drugs. Results: In patients with JME, GABA and NAA were reduced in the thalamus (p = 0.03 and p = 0.02), whereas frontal GABA and glutamine were elevated (p = 0.046 and p = 0.03). MRS revealed reduced NAA in the thalamic gray matter contralateral to the handedness (p = 0.04 each). These changes were found consistently in patients treated with new antiepileptic drugs and with valproate, although the extent of metabolic changes differed between these treatments. Significance: Decreased thalamic and increased frontal GABA suggest a dysfunction of GABAergic neurotransmission in these brain regions of patients with JME. The NAA decrease in the gray matter of the thalamus may hint to a damage of GABAergic neurons, whereas frontal increase of GABA and its precursor glutamine may reflect increased density in GABAergic neurons due to subtle cortical disorganization in the thalamofrontal network. KEY WORDS: Juvenile myoclonic epilepsy, GABA, MR spectroscopy, Thalamofrontal.
Accepted April 12, 2014; Early View publication June 5, 2014. *Department of Neuroradiology, Goethe University Frankfurt, Frankfurt, Germany; †Centrum of Epilepsy Hessen, Philipps-Universit€at Marburg, Marburg, Germany; and ‡Department of Paediatric Neurology, Goethe University Frankfurt, Frankfurt, Germany Address correspondence to Elke Hattingen, Department of Neuroradiology, Goethe University Frankfurt, Schleusenweg 2-16, Frankfurt 60528, Germany. E-mail: [email protected] Wiley Periodicals, Inc. © 2014 International League Against Epilepsy
Juvenile myoclonic epilepsy (JME) is characterized by awakening myoclonus, often associated with tonic–clonic seizures, absences, and neuropsychological frontal lobe pathology.1,2 Abnormal oscillations of the thalamocortical networks are considered to trigger myoclonic seizures in JME.3 The thalamus amplifies and synchronizes thalamocortical rhythms, which involve pyramidal cells of the motor cortex.4 This thalamocortical network is reciprocally connected by excitatory glutamatergic and inhibitory c-aminobutyric acid (GABA)ergic projections.5 Congenital
1030
1031 Frontal and Thalamic GABA Concentration in JME alterations of GABAA receptors play a key role in JME, and changes in GABAA receptor were found in positron emission tomography (PET) imaging.6,7 Apart from the receptor level, there is increasing evidence that the network between the frontal lobe and the thalamus is structurally disturbed during early brain development in patients with JME.8 GABA is not only functionally but also metabolically linked with its excitatory counterpart glutamate,9 as glutamine is the precursor of both. GABA and glutamate, once released from neurons, are taken up into astrocytes, which convert glutamate into glutamine. Impairments of these glial-neuronal interactions result in seizures.10 High-field proton magnetic resonance spectroscopy (1H MRS) allows noninvasive quantification of these neurotransmitters. However, glutamine and glutamate are not easily quantifiable at field strengths below 4 Tesla when their overlapping peaks start to separate and GABA measurement requires additional editing. In contrast, the neuronal marker N-acetyl aspartate-glutamate (NAA) is quantifiable even at lower field strengths, and reduced NAA was observed in brain structures that are involved in epilepsy.11–15 Only few studies measured in vivo concentrations of GABA or glutamate and glutamine separately (Glx as sum of both, Glx) in patients with JME.16–18 Petroff et al.16,17 found decreased GABA concentrations in the occipital brain of JME patients, whereas Simister et al.14,18 reported increased GABA levels in the occipital lobe of 15 patients with idiopathic generalized epilepsy (IGE), including two patients with JME. Levels of Glx were increased in the frontal lobe, the insula, and striatum of patients with JME, raising the question whether this increase refers to glutamine or glutamate or both. Herein we present data on metabolite concentrations in brain structures that are most relevant for JME (thalamus, frontal lobe, and motor cortex) focusing on changes of GABA, glutamate, glutamine, and NAA. In addition to 1H MRS with short echo time (TE) to measure glutamate and glutamate separately, a specific MR spectroscopic GABA A
C
editing sequence was employed. Because metabolite concentrations can also be changed by the medication used for treatment,19–24 the study was also designed to obtain information of those effects, as well hemispheric and tissue differences.
Materials and Methods Study population This prospective noninterventional study was approved by the institutional review board (Ethics committee at the University Hospital Marburg, 150/11), and written informed consent was obtained from each participant prior to inclusion. Patients were enrolled by an epilepsy center, and diagnosis of JME was confirmed by electroencephalography (EEG) and semiology. Diagnosis of JME was made according to the criteria of the 2012 International Consensus Criteria by an epilepsy expert,25 defined by myoclonic jerks without loss of consciousness on awakening, normal interictal EEG with ictal high-amplitude generalized epileptiform discharges (class I), and otherwise normal intelligence and normal neurologic examination. Patients and healthy age-matched controls had a normal morphologic cranial magnetic resonance imaging (MRI) study. Sixteen JME patients and 18 healthy subjects were enrolled between 2011 and 2012. None of the patients had a seizure 24 h before the MR examination. MR spectroscopy All spectroscopic examinations were performed at 3 Tesla. The entire protocol, which was initially designed to obtain 1H- and 31P-MRS data from brain areas with potential pathologic relevance for JME, was divided into two sessions (MR parameter details are indicated in the supporting information): 1 Single-voxel spectroscopy (SVS) to measure 1H MRS detectable metabolite concentrations in the thalamus, the frontal lobe, and the motor cortex of the hemisphere D
B
Figure 1. Single voxel position in the frontal lobe (A) and thalamus (B), and representative spectra for analysis of short echo time (TE) (C) and GABA-edited (D) signals. The dotted red line in C shows the result from the LCModel fit. Intensity adjusted basis spectra for MI (blue) and Gln (orange), as obtained by the fit, are also included. GABA-editing by MEGA-PRESS (TE = 70 msec) is shown in (D). The black line represents the difference between the blue (editing pulse at 4.1 ppm) and green signal (editing pulse at 1.9 ppm). The fitting procedure quantifies the GABA signal at 3 ppm (dotted red line). Epilepsia ILAE Epilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
1032 E. Hattingen et al. contralateral to the handedness (Fig. 1). The SVS data included a short TE (30 msec) point resolved selective spectroscopy (PRESS) sequence to measure the main 1H MRS detectable metabolites (see below) and a MescherGarwood (MEGA)-PRESS sequence at a TE of 70 msec, dedicated to the detection of GABA by using a a pair of selective excitation pulses at 1.9 ppm.27 2 In a separate study short TE (30 msec) 1H MR spectroscopic imaging (MRSI) was performed. A weighted circular phase-encoding scheme was used with a 24 9 24 matrix and 240 9 240 mm2 field of view. The volume of interest was selected by a combination of pointresolved selective spectroscopy and outer volume suppression. Before Fourier transformation, the matrix was extrapolated to 48 9 48, resulting in a 5 9 5 mm2 inplane grid size. To obtain metabolite concentrations in gray matter (GM) or white matter (WM), co-registration with segmented anatomic images for GM and WM was performed to allow manual selection of the small MRSI voxels with only GM and WM (see Figure S1 in Supporting Information). The slices were coronal oriented to measure brain metabolites in GM and WM of both hemispheres including the thalamus and the motor cortex. Data processing MRS data Short TE SVS and MRSI spectra were analyzed quantifying the main metabolites NAA (N-acetyl-aspartate/N-acetylaspartate-glutamate), Cr (creatine/phosphocreatine), Cho (choline-containing compounds), and MI (myoinositol). The larger SVS voxel > 8 ml together with high spectral resolution at 3 T allowed quantification of glutamine and glutamate separately, whereas Glx as sum of glutamine and glutamate was quantified for the much smaller MRSI voxels (0.30 ml). No GABA concentrations were available from MRSI spectra, since detection and quantification of GABA requires special sequences (e.g. MEGA-PRESS).16–18,26 The GABA-edited data (Fig. 1D) from MEGA-PRESS spectra were analyzed with the tool jMRUI while the shortTE 1H MRS spectra were analyzed with the software tool LCModel (http://s-provencher.com). Partial volume for GM, WM, and cerebrospinal fluid (CSF) was obtained from the segmented anatomic images. Further details of data acquisition and processing are provided in the Supporting Information. Statistics Statistical analysis was performed with STATISTICA (version 7.1; StatSoft, Tulsa, OK, U.S.A.). For all tests, p < 0.05 was considered to be statistically significant. Comparing JME patients to controls Concentration differences for GABA and NAA between patients and control subjects were tested with analysis of Epilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
variance (ANOVA) using the coil loading as covariate. Different target regions were considered as repetitive measurements. Contrast analysis in ANOVA was performed for each metabolite and target region. Effects of antiepileptic drugs Previous data revealed that antiepileptic drugs (AED) have an effect on brain metabolites in 1H MRS. It has been shown that valproate induces changes in cerebral myoinositol and glutamine concentrations, whereas new AEDs may increase GABA levels in patients with epilepsy.21–25,29 Therefore, we divided the patients into two groups: one group included patients with valproate treatment, whereas the other group covered all patients who received any medication but no valproate. Differences in myoinositol and glutamine between these two groups were tested using ANOVA.
Results Study population For three controls and one class I JME patient treated with lamotrigine, no MRSI was performed due to subject’s cancellation. These subjects were not included in the analysis. The remaining 30 subjects had a median age for patients and controls of 25 years. Age of the patients was 16–36 years, except of one 54-year-old patient; age range of controls was 16–38 years. The control group consisted of seven females and eight males, whereas the JME group included 10 females and five males. Clinical data of the JME patients are given in Table 1. Typical semiology with myoclonic jerks occurring on or after awakening without loss of consciousness was reported by all patients. Thirteen of the 15 patients had typical myoclonic seizures as first symptom of JME, which manifested at the age of 10–19 years, whereas in one patient, myoclonic jerks occurred together with absences and follow-up EEG was typical for juvenile absence epilepsy (Table 1). In another patient the myoclonus was associated with awakening tonic–clonic seizure early after onset, fulfilling the class 2 criteria according the International Consensus.25 The remaining 13 patients with JME semiology and typical EEG findings fulfilled the class 1 criteria according the International Consensus.25 Seven patients had generalized tonic– clonic seizures and absence seizures over the years of illness. Two patients had generalized tonic–clonic and two patients had absence seizures. All but one patient had antiepileptic therapy, which consisted of valproate, levetiracetam, lamotrigine, topiramate, or the combination of those. To ensure that the uncertain case of JME or gender differences did not influence the major effects of neurotransmitter changes, we performed a second statistical analysis on all SVS data, only, matching the groups for age and gender (10 female and five male, mean age 26) and replacing the data of the uncertain case by those from the other class I JME patient.
1033 Frontal and Thalamic GABA Concentration in JME Table 1. Clinical characteristics of the patients and antiepileptic drugs Patient
Age (y), gender
Age at onset (y)
Disease duration (y)
JME class according26
Seizure characteristics
AEDs
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
35, m 26, f 16, f 36, m 31, f 25, m 25, m 16, m 32, f 20, f 16, f 54, f 17, f 21, f 20, f
17 18 13 18 11 11 19 14 13 13 10 10 11 16 16
18 8 3 18 20 14 6 3 19 7 6 44 6 5 4
1 1 1 1 1 1 1 1 1 1 1 1
M, GTC, A M M M, GTC, A M, GTC, A M, GTC M M, GTC, A M, GTC, A M, A M, GTC, A M M, A M, GTC M, GTC, A
VPA, LTG LTG VPA LEV LEV VPA VPA VPA VPA LEV LEV, LTG None LEV VPA, LTG VPA, TPM
a
1 2
Y, years; m, male, f, female; AED, antiepileptic drugs; M, myoclonia; GTC, generalized tonic–clonic seizure; A, absence; VPT, valproate; LEV, levetiracetam; LTG, lamotrigine; TPM, topiramate. a Included as JME. EEG showed generalized 3/s spike-slow wave discharges intensified by photo stimulation characterizing juvenile absence epilepsy.
Metabolite concentrations Concentrations of the metabolites GABA, NAA, and glutamine as obtained from SVS magnetic resonance spectroscopy (MRS) of the hemisphere contralateral to the handedness differed for the selected brain areas between healthy control subjects and patients (Fig. 2) showing: 1 Increase in GABA concentrations in the frontal region (p = 0.046) and decreased GABA concentrations in the thalamus (p = 0.03). 2 Decrease in NAA concentrations in the thalamus (p = 0.02). 3 Increase of glutamine concentration in the frontal area (p = 0.03). These results were confirmed by the additional analysis of the second cohort with exactly matched gender and age. Table 2 also shows metabolites of the patient group dichotomized into a group treated with valproate and a group without valproate treatment. These data should exclude that above-mentioned significant metabolite changes reflect an effect of AED instead of being disease related. Although the statistical power of a comparison between each patient group and controls is rather low, the above-mentioned significant changes in JME patients were found equally in each subgroup. The previously reported effects of valproate on metabolites were also found in terms of a consistent increase in glutamine and a decrease in myoinositol when comparing the two patient groups, and ANOVA revealed significance (p = 0.045); contrast analysis in ANOVA confirmed increased glutamine in the thalamus. Although the SVS data provided only metabolite concentration changes in the target region of the hemisphere contralateral to the handedness, the MRSI data also included
information for the target region ipsilateral to the handedness and allowed for discrimination between GM and WM (Supporting Information). However, concentrations of GABA and glutamine were not obtained with these data, and the frontal region was not covered by the coronal slice. An intra-individual comparison for control subjects showed significantly higher NAA concentration in the GM of the thalamus contralateral to the handedness compared to ipsilateral (p = 0.006). Comparing patients to controls revealed significantly decreased concentrations of NAA (p = 0.04) and myoinositol (p = 0.002) for this tissue, confirming the SVS data (Appendix S1). Both changes were also visible in the hemisphere ipsilateral to the handedness, but they did not reach significance. Furthermore, choline was significantly decreased in the thalamic GM of the hemisphere contralateral to the handedness (p = 0.04). Dichotomizing into the two groups with different medication shows the most prominent decrease of myoinositol and NAA in the thalamic GM of patients treated with valproate.
Discussion This study focuses on potentially affected brain areas when measuring GABA, glutamine, and glutamate concentration in JME patients. The primary findings were reduced concentrations of neuronal metabolites GABA and NAA in the thalamus, whereas GABA and glutamine were increased in the frontal lobe. Although these results are based on SVS data targeting the contralateral to handedness side, the MRSI data confirmed reduced NAA concentrations in the thalamus contralateral to the handedness in JME patients, also pointing to more prominent changes in the gray matter. Petroff et al.16 found a decrease of GABA concentration in Epilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
1034 E. Hattingen et al.
Figure 2. Metabolites from the glutamate/GABA-glutamine cycle and the neuronal marker NAA as obtained from SVS of contralateral to the handedness hemisphere for healthy subjects (Prob) and patients (Pat). The box plots represent mean values with standard errors; standard deviation is marked by whiskers. Epilepsia ILAE Epilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
the occipital brain region of JME patients, but they did not make an attempt to localize GABA changes in different brain areas. Nevertheless, these occipital measurements may have included thalamic regions, since SVS was performed with an 8-cm surface coil and a large voxel size of 14 ml. In contrast, Simister et al.17,18 used localized PRESS with double quantum filter to measure GABA in the frontal and occipital lobe of patients with IGE, detecting GABA increase only in the occipital area. Only part of these IGE patients had the diagnosis of JME (2 of 15 patients for occipital measurements and 7 of 21 patients for frontal measurements). In contrast, our study included JME patients meeting the internationally proposed class I (n = 14) or class II (n = 1) JME criteria.25 Apart from the methodologic differences, these results suggest that regional differences of altered GABA concentration are responsible for different phenotypes of IGE seizures. In a PET study investigating a selective antagonist of GABAA receptors, [11C]-flumazenil (FMZ), Koepp et al.7 showed an increased binding in the cerebral cortex and thalamus of 10 IGE patients; in 5 JME patients there was an accentuation of the dorsolateral prefrontal areas when compared to healthy controls. Because a mutation in the a1 subunit of the GABAA receptor was found in a family with autosomal JME,6 one key in the pathogenesis of JME was attributed to the receptor level. In vitro studies revealed that the majority of the mutations result in a reduction of GABA-activated chloride currents in recombinant receptors.27 GABA is the main inhibitory transmitter in the adult brain; therefore, it was suggested that the lack of GABA-activated chlorine currents caused abnormal excitability of cortical neurons, thereby resulting in myoclonic or general seizures. Apart from loss of function, there is increasing evidence that epilepsy-causing mutations in GABAA receptors might also be associated with abnormal development of neuronal networks, which is anticipated as one of the critical mechanisms leading to the disease. During brain development, GABA seems to act as a excitatory neurotransmitter, playing a key role in various aspects of the maturation process, such as neuronal migration.28 Furthermore, it has been found recently that other mutations associated with JME are deeply involved in cerebral corticogenesis: The myoclonin1/EFHC1 microtubule-associated protein is involved in cell division and radial migration during cortical development, and mutations in the EFHC1 gene caused a marked disturbance of cortical development in vivo.8 BRD2, the locus for another JME-inducing gene (EJM1), is a putative transcriptional regulator from a gene family that is expressed during cerebral development.5 Abnormal brain development in JME is supported by structural MRI studies. Voxel-based volumetry and diffusion tensor imaging reveal changes predominantly in the prefrontal cortex involving the gray and underlying white matter.29–31 Furthermore, the connectivity of different brain regions was found to be
1035 Frontal and Thalamic GABA Concentration in JME Table 2. Values with standard deviation (italic) of all metabolites measured with single voxel spectroscopy (SVS). Significant differences between patients and control subjects are marked in bold. Values represent concentrations in mM Patients Region Motor area
Frontal
Thalamus
Control
Metabolite Cr MI Cho NAA Glu Gln GABA Cr MI Cho NAA Glu Gln GABA Cr MI Cho NAA Glu Gln GABA
7.09 4.14 1.30 11.10 7.06 4.21 1.97 5.56 3.51 1.36 8.79 5.25 2.20 3.09 6.44 3.50 1.53 9.56 5.66 2.67 4.29
Total 0.86 0.79 0.17 1.57 1.43 2.05 0.46 1.27 0.43 0.31 1.39 1.33 1.29 0.63 0.76 0.54 0.16 1.02 0.76 1.17 0.50
7.07 3.83 1.39 10.64 6.74 4.17 2.04 5.60 3.31 1.42 8.71 5.24 3.16 3.56 6.39 3.52 1.55 8.95 5.39 3.17 3.85
affected, either by means of a decrease to frontal areas or by an increase to the motor pathways.32 Although all studies reported structural changes in patients with JME, the areas and the direction of changes still differed between these studies.29,30,33 Increased frontal concentration levels of GABA and of the neurotransmitter precursor glutamine found in this study corroborate the findings of an involvement of the frontal cortex. MRS measures the total amount of GABA that is available. Thus, the GABA concentration mainly reflects the fractional volume of GABAergic neurons adjusted by the vesicular and synaptic GABA content of those neurons. Furthermore, metabolic concentrations measured by MRS do not simply represent the neurotransmitter pool, but also reflect the metabolism. Increased GABAA receptor binding as observed for the JME patients in the PET study described previously,7 would be in line with an increased density in GABAergic neurons in the frontal lobe. An increased density in GABAergic neurons also supports the hypothesis that developmental disorganization of the cortex in form of cortical and subcortical dystopic neurons (“cortical microdysgenesis”) may play a key role in JME.7,34 This would also explain why higher frontal gray matter concentration of JME patients were found compared to controls.29,35 The relevance of glutamine (maybe plus glutamate) for JME is supported by previous findings. Simister et al.17,18 found increased levels of Glx (glutamate + glutamine) in the frontal and occipital lobe of patients with IGE. In accordance with our study results, frontal Glx increase was also found
Valproate 1.03 0.76 0.28 1.43 1.14 1.57 0.68 1.21 0.68 0.28 1.34 0.80 1.11 0.64 0.94 0.76 0.23 1.26 1.01 0.80 0.67
6.83 3.54 1.27 10.00 6.35 4.91 1.89 5.17 3.02 1.34 8.01 4.90 3.28 3.42 6.09 3.13 1.47 8.37 5.10 3.54 3.63
Nonvalproate 1.02 0.59 0.34 1.43 0.92 1.78 0.70 1.30 0.48 0.29 1.10 0.66 1.33 0.56 1.06 0.78 0.27 1.35 1.13 0.90 0.71
7.36 4.17 1.51 11.35 7.18 3.31 2.20 6.10 3.65 1.51 9.50 5.64 3.04 3.73 6.72 3.96 1.65 9.61 5.74 2.77 4.10
1.06 0.83 0.12 1.13 1.26 0.68 0.67 0.94 0.78 0.23 1.21 0.79 0.88 0.55 0.71 0.47 0.11 0.78 0.79 0.44 0.58
in the subgroup of seven JME patients.18 Another study found a frontal Glx/Cr increase together with NAA/Cr decrease in JME patients with personality disorders, supporting the hypothesis that typical personality changes of JME patients are related to the frontal lobe dysfunction.15 The decrease of GABA and NAA in the thalamus may be attributed to a reduction of the GABAergic neurons. NAA is formed in the neuronal mitochondria, and hereby the NAA decrease marks neuronal dysfunction or neuronal loss. NAA decrease in the thalamus of JME patients has been reported previously.12–14 Inconsistent results were found in voxel-based morphometry studies concerning neuronal damage in the thalamus of JME patients.30,35,36 However, some studies hint to a regional abnormality in the anterior and medial parts of the thalamus, which are connected to the frontal lobe, supporting the role of thalamofrontal network dysfunction in the epileptogenesis of JME. In our study, concentration decreases of NAA, but also of Cho and MI were most prominent in the GM of the thalamus contralateral to the handiness, reflecting a higher vulnerability of this tissue in JME. Language specialization of the thalamus contralateral to the handedness and functional demand according to the handedness may enhance metabolic changes in the thalamus.37 The hemispheric difference can also explain why Savic et al.38 did not find a significant NAA decrease in the right thalamus, which is ipsilateral to the handedness in the majority of humans. However, it remains unclear whether these changes are of secondEpilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
1036 E. Hattingen et al. ary degenerative nature or if they are part of a cerebral development failure. An experimental study in heterozygous BRD2 mice hints to a development failure.5 It is worth mentioning that NAA concentration in the thalamus contralateral to the handedness is higher in healthy subjects (see Appendix S1).39 All JME patients except one were receiving antiepileptic treatment. Because interruption of this treatment is contraindicated, potential drug effects on brain metabolite concentrations have to be considered. In particular, it should be excluded that the above-mentioned changes in GABA, glutamine, and NAA are drug-induced. According to previously published data, valproate should not change the GABA and NAA concentrations, whereas other medications might increase GABA.19,21–24 On the other hand, valproate increases cerebral concentrations of glutamine and decreases the myoinositol concentration.22,23 However, dichotomized patient data reveal that both, thalamic GABA decrease as well as frontal increase of GABA and glutamine, should be related to the disease, given that these changes were found equally in both groups. Of note, the increased concentration of neuronal metabolites NAA, GABA, and glutamate in the frontal area and the motor cortex were more prominent under therapy with newer AED compared with the valproate treatment, which might reflect a direct effect on cerebral neuronal metabolism for these drugs.21,24 Glutamate and GABA are interconnected via the glutamate/ GABA-glutamine cycle and NAA is discussed as a reservoir for glutamate.9,40 These findings also support the hypothesis that some drugs interacting with neurotransmitter systems may have a neurotropic effect.20 In conclusion, changes of GABA concentration in the thalamus and frontal lobe as key regions for seizure triggering may be essential in JME. The decrease of the neuronal marker NAA in the gray matter of the thalamus implicates a damage of GABAergic neurons. The increase of GABA and its precursor glutamine hints to inborn errors of the thalamofrontal neuronal network, either on the receptor level or in the form of failures in corticogenesis. Although GABA concentrations are affected by some AEDs, we can exclude a simple treatment effect. Larger MRS trials should approve the generalizability of our study results by analyzing neurotransmitter levels in JME patients before and after initiating AED treatment.
Funding There was no external funding for the study.
Disclosure All authors declare that none of the authors has any conflict of interest to disclose. All authors confirm that they have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines. Epilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
References 1. Canevini MP, Mai R, Di Marco C, et al. Juvenile myoclonic epilepsy of Janz: clinical observations in 60 patients. Seizure 1991;1:291–298. 2. Devinsky O, Gershengorn J, Brown E, et al. Frontal functions in juvenile myoclonic epilepsy. Neuropsychiatry Neuropsychol Behav Neurol 1997;10:243–246. 3. O’Muircheartaigh J, Vollmar C, Barker GJ, et al. Abnormal thalamocortical structural and functional connectivity in juvenile myoclonic epilepsy. Brain 2012;135:3635–3644. 4. Gotman J, Grova C, Bagshaw A, et al. Generalized epileptic discharges show thalamocortical activation and suspension of the default state of the brain. Proc Natl Acad Sci USA 2005;102:15236– 15240. 5. Velısek L, Shang E, Velıskova J, et al. GABAergic neuron deficit as an idiopathic generalized epilepsy mechanism: the role of BRD2 haploinsufficiency in juvenile myoclonic epilepsy. PLoS ONE 2011;6: e23656. 6. Cossette P, Liu L, Brisebois K, et al. Mutation of GABRA1 in an autosomal dominant form of juvenile myoclonic epilepsy. Nat Genet 2002;31:184–189. 7. Koepp MJ. Juvenile myoclonic epilepsy– a generalized epilepsy syndrome? Acta Neurol Scand Suppl 2005;181:57–62. 8. de Nijs L, Leon C, Nguyen L, et al. EFHC1 interacts with microtubules to regulate cell division and cortical development. Nat Neurosci 2009;12:1266–1274. 9. Bak LK, Schousboe A, Waagepetersen HS. The glutamate/GABAglutamine cycle: aspects of transport, neurotransmitter homeostasis and ammonia transfer. J Neurochem 2006;98:641–653. 10. Eid T, Ghosh A, Wang Y, et al. Recurrent seizures and brain pathology after inhibition of glutamine synthetase in the hippocampus in rats. Brain 2008;131:2061–2070. 11. Savic I, Lekvall A, Greitz D, et al. MR spectroscopy shows reduced frontal lobe concentrations of N-acetyl aspartate in patients with juvenile myoclonic epilepsy. Epilepsia 2000;41:290–296. 12. Mory S, Li L, Guerreiro C, et al. Thalamic dysfunction in juvenile myoclonic epilepsy: a proton MRS study. Epilepsia 2003;44:1402– 1405. 13. Haki C, G€um€ustas OG, Bora I, et al. Proton magnetic resonance spectroscopy study of bilateral thalamus in juvenile myoclonic epilepsy. Seizure 2007;16:287–295. 14. Lin K, Carrete H Jr, Lin J, et al. Magnetic resonance spectroscopy reveals an epileptic network in juvenile myoclonic epilepsy. Epilepsia 2009;50:1191–1200. 15. de Araujo Filho GM, Lin K, Lin J, et al. Are personality traits of juvenile myoclonic epilepsy related to frontal lobe dysfunctions? A proton MRS study. Epilepsia 2009;50:1201–1209. 16. Petroff OA, Hyder F, Rothman DL, et al. Homocarnosine and seizure control in juvenile myoclonic epilepsy and complex partial seizures. Neurology 2001;56:709–715. 17. Simister RJ, McLean MA, Barker GJ, et al. A proton magnetic resonance spectroscopy study of metabolites in the occipital lobes in epilepsy. Epilepsia 2003;44:550–558. 18. Simister RJ, McLean MA, Barker GJ, et al. Proton MRS reveals frontal lobe metabolite abnormalities in idiopathic generalized epilepsy. Neurology 2003;61:897–902. 19. Petroff OA, Rothman DL, Behar KL, et al. Effects of Valproate and other antiepileptic drugs on brain glutamate, glutamine, and GABA in patients with refractory complex partial seizures. Seizure 1999;8:120– 127. 20. Moore GJ, Bebchuk JM, Wilds IB, et al. Lithium-induced increase in human brain grey matter. Lancet. 2000 Oct 7;356(9237):1241–1242. Erratum in: Lancet 2000;356:2104. 21. Kuzniecky R, Ho S, Pan J, et al. Modulation of cerebral GABA by topiramate, lamotrigine, and gabapentin in healthy adults. Neurology 2002;58:368–372. 22. Simister RJ, McLean MA, Barker GJ, et al. The effect of sodium valproate on proton MRS visible neurochemical concentrations. Epilepsy Res 2007;74:215–219. 23. Garcia M, Huppertz HJ, Ziyeh S, et al. Valproate-induced metabolic changes in patients with epilepsy: assessment with H-MRS. Epilepsia 2009;50:486–492.
1037 Frontal and Thalamic GABA Concentration in JME 24. Doelken MT, Hammen T, Bogner W, et al. Alterations of intracerebral c-aminobutyric acid (GABA) levels by titration with levetiracetam in patients with focal epilepsies. Epilepsia 2010;51:1477–1482. 25. Kasteleijn-Nolst Trenite DG, Schmitz B, Janz D, et al. Consensus on diagnosis and management of JME: from founder’s observations to current trends. Epilepsy Behav 2013;28(Suppl. 1):87–90. 26. Mescher M, Merkle H, Kirsch J, et al. Simultaneous in vivo spectral editing and water suppression. NMR Biomed 1998;11:266–272. 27. Macdonald RL, Gallagher MJ, Feng HJ, et al. GABA(A) receptor epilepsy mutations. Biochem Pharmacol 2004;68:1497–1506. Review. 28. Ben-Ari Y, Khalilov I, Kahle KT, et al. The GABA excitatory/ inhibitory shift in brain maturation and neurological disorders. Neuroscientist 2012;18:467–486. 29. Woermann FG, Free SL, Koepp MJ, et al. Abnormal cerebral structure in juvenile myoclonic epilepsy demonstrated with voxel-based analysis of MRI. Brain 1999;122:2101–2108. 30. Kim JH, Lee JK, Koh SB, et al. Regional grey matter abnormalities in juvenile myoclonic epilepsy: a voxel-based morphometry study. Neuroimage 2007;s37:1132–7113. 31. Deppe M, Kellinghaus C, Duning T, et al. Nerve fiber impairment of anterior thalamocortical circuitry in juvenile myoclonic epilepsy. Neurology 2008;71:1981–1985. 32. Vollmar C, O’Muircheartaigh J, Barker GJ, et al. Motor system hyperconnectivity in juvenile myoclonic epilepsy: a cognitive functional magnetic resonance imaging study. Brain 2011;134:1710– 1719. 33. Liu M, Concha L, Beaulieu C, et al. Distinct white matter abnormalities in different idiopathic generalized epilepsy syndromes. Epilepsia 2011;52:2267–2275.
34. Meencke HJ, Janz D. Neuropathological findings in primary generalized epilepsy: a study of eight cases. Epilepsia 1984;25:8–21. 35. Mory SB, Betting LE, Fernandes PT, et al. Structural abnormalities of the thalamus in juvenile myoclonic epilepsy. Epilepsy Behav 2011;21:407–411. 36. Saini J, Sinha S, Bagepally BS, et al. Subcortical structural abnormalities in juvenile myoclonic epilepsy (JME): MR volumetry and vertex based analysis. Seizure 2013;22:230–235. 37. Eidelberg D, Galaburda AM. Symmetry and asymmetry in the human posterior thalamus. I. Cytoarchitectonic analysis in normal persons. Arch Neurol 1982;39:325–332. 38. Savic I, Ostermann Y, Helms G. MRS shows syndrome differentiated metabolite changes in human-generalized epilepsies. Neuroimage 2004;21:163–172. 39. Nagae-Poetscher LM, Bonekamp D, Barker PB, et al. Asymmetry and gender effect in functionally lateralized cortical regions: a proton MRS imaging study. J Magn Reson Imaging 2004;19:27–33. 40. Clark JF, Doepke A, Filosa JA, et al. N-acetylaspartate as a reservoir for glutamate. Med Hypotheses 2006;67:506–512.
Supporting Information Additional Supporting Information may be found in the online version of this article: Appendix S1. 1H MRSI and SVS in JME.
Epilepsia, 55(7):1030–1037, 2014 doi: 10.1111/epi.12656
